what technology made it into gffx from 3dfx?
- Thread starter skilzygw
- Start date
Deflection
Newcomer
An extra 3 months?
Edit: "Time Dialation" would have been more appropriate
In all seriousness though, the pace the engineering teams at Nvidia, ATI, and 3DLabs have graphics on is quite amazing!
Edit: "Time Dialation" would have been more appropriate
In all seriousness though, the pace the engineering teams at Nvidia, ATI, and 3DLabs have graphics on is quite amazing!
I agree, the delays are what makes the NV30 a very special piece of hardware
Without 3dfx's contributions, it would not have been possible.
Seriously tho, it is the lack of 3dfx features that seems very apparent. At least one would expect it to have the same AA implementation(albeit with optimizations), but it is conspicuously absent.
Without 3dfx's contributions, it would not have been possible.
Seriously tho, it is the lack of 3dfx features that seems very apparent. At least one would expect it to have the same AA implementation(albeit with optimizations), but it is conspicuously absent.
how about new ideas on how to implement DX9 that are not obviously 3dfx as they werent thought of back then!
Everyone seems to think 3dfx tech means there would have been jittereed RGSS, Rampage specs which have been surpassed anyway and Gigapixel defferred rendering. Lucky it didnt mean slow trilinear eh
Persoannly I reckon its the molex connector
Everyone seems to think 3dfx tech means there would have been jittereed RGSS, Rampage specs which have been surpassed anyway and Gigapixel defferred rendering. Lucky it didnt mean slow trilinear eh
Persoannly I reckon its the molex connector
megadrive0088
Regular
"I agree, the delays are what makes the NV30 a very special piece of hardware.
Without 3dfx's contributions, it would not have been possible."
ROFLMAO!
Without 3dfx's contributions, it would not have been possible."
ROFLMAO!
Funny how people think that "technology" is some bunch of blueprints that can be bought and just made use of.
I think NVidia explained it clearly enough: 3dfx's engineers insight, talent, and ideas contributed to NV30 ( and even on GF4 already .. some video processing stuff and AA filtering on scanout )
Its not like "hm lets see.. we take rampage's TMU's .. our crossbar mem controller and Gigapixel's tiling system and throw everything into one killer chip"
I think NVidia explained it clearly enough: 3dfx's engineers insight, talent, and ideas contributed to NV30 ( and even on GF4 already .. some video processing stuff and AA filtering on scanout )
Its not like "hm lets see.. we take rampage's TMU's .. our crossbar mem controller and Gigapixel's tiling system and throw everything into one killer chip"
megadrive0088
Regular
I think that's how many people see it, actually. Admittedly, that's how I used to think after Nvidia acquired 3Dfx and thus Rampage+GigaPixel.
This was considerably more than a paper launch in 3dfx terms. When VSA-100 was announced at Comdex it seems as though theyr weren't even close to tape-ou at that point, so there had zero to show for it. At least here we had fully operable chips / boards being shown running full NV30 OpenGL demos and new game titles.
megadrive0088 said:
its funny, because that really is how people see it.
whenever i see someone type "3dfx tech" or "artx tech" i cringe.
the same goes to anyone who thinks ATI is sitting there right now pondering what to do with the R350 in order to "counter" the geforcefx (if that even exists, my guess is that they will just launch an R400 this fall with an RV350, like they've done with every other R product launch in their history). Whatever goes into any chip is practicly set in stone about 1 to 1.5 years before we even hear about it.
Certainly 1-1.5 years before it hits the stores, but we hear about it much sooner. Hell, the NV30 has been in rumors for the past 2 years alone.
I'll agree with no_way. NV25's Accuview had already contributions from former 3dfx tech. I have my suspicions that the mentioned in a diagram at digit life Tile HSR Logic is in fact NOT completely irrelevant to Gigapixel at the other. Could be wrong though.
RussSchultz said:
true, but most of those are just that, rumors. often they are not based on any factual leaked evidence, they're simply projections of what can be done, based on speculation on the future state of the industry. we know DDRII is going to be available at such and such time, thus we infer that it is nVidias or ATIs best interest to be using whatever is the best available. It doesnt take a genius to realize that they've known this as well, and are taking it into account. They're probably all enthusiasts just like us. it cracks me up that certain people (wont name names; spy wannabees) will pass all this wild speculation as if its some sort of unfolding drama, when in reality its a very slow and arduous process to bring this silicon to the market. its also humorous that when there are uncertainties, like with whatever patents or designs that 3dfx had been working on before their acquisition, that some lable it as "tech" and figure nV is just gonna slap it on to their existing design, as if its some car part like a turbo, that you can just slide under the hood. Im certainly no engineer, but I am out of elementry school, thus its really not a stretch for me, or anyone who's passed 6th grade, to understand how complex a 125m transistor part really is, from the math that spawns it, all the way up to its physical implementation on silicon.
I can understand that they bought the engineers over and that helps your team with ideas, experience etc... But then why bother aquiring 3dfx and gigapixel properties? Just hire the damn people after their company has gone bust. It must have been that they saw some technology they liked and wanted to aquire it.
It's all about patents. If for example ATI bought all the 3dfx and Gigapixel patents, while Nvidia picked up the 3dfx engineers, then anything that the 3dfx engineers might bring to Nvidia might be a fast track to a law suit.
